labourdemand_wake
Description
This document describes the variables used in the paper “Labour demand in the wake of a shock: A dose–response approach”. The main variable of interest is labour demand, measured as the logarithm of the number of vacancies from the Lightcast dataset. Vacancies are classified into total, essential, and non-essential sectors, based on ATECO codes and in line with the Italian government’s decree of 22 March 2020. In addition, the dataset includes labour market indicators, mortality and contagion measures, as well as demographic characteristics at the province (NUTS-3) level in Italy during the COVID-19 pandemic. This document provides an overview of the variables included, their type, and their ranges. The vacancy data from the Lightcast dataset is confidential; therefore, we provide a synthetic dataset that reproduces the structure of the original data.
Files
Steps to reproduce
To estimate the effects, we employ the Stata module ctreatreg, developed by Cerulli (2015) , which is designed for estimating Dose–Response Treatment Models under (continuous) treatment endogeneity and heterogeneous responses to observable confounders. The program assigns different “levels” of treatment (or dose $t$) to treated units, ranging from $0$ (absence of treatment) to $100$ (maximum treatment level). In this context, the parameter of interest is the Dose–Response Function (DRF) of $y$ (labour demand-vacancies) on $t$ (contagions). Our study uses COVID-19 contagion rates as the treatment variable, designating “low-treated” units as the control group. We transformed the treatment variable on a scale from 0 to 100 to facilitate the interpretation (we refer to them as “levels”). The ctreatreg command estimates this DRF, which corresponds to the “Average Treatment Effect (ATE) at treatment level $t$” , along with other causal parameters of interest such as the overall ATE, the Average Treatment Effect on the Treated (ATET), and the Average Treatment Effect on the Untreated (ATENT).
Institutions
- Gran Sasso Science InstituteAbruzzo, L'aquila